U.S. DOJ Considers Updating Antitrust Guidelines on AI Safety
nashnova research
The U.S. Department of Justice is weighing whether to extend cybersecurity antitrust guidelines to cover AI, but no AI company has actually contacted the DOJ — and the FTC has already pushed back hard. The split between the two enforcers means a unified green light is unlikely any time soon.
What exactly is the DOJ considering?
Current cybersecurity antitrust guidelines date to the Obama era. They let companies share threat intelligence without triggering anti-competition rules.
Deputy Attorney General Stanley Woodward said the government is evaluating whether to extend that framework to AI safety.
This means → if updated, frontier AI firms could collaborate more deeply on safety risks without fear of being accused of collusion.
Who is pushing for this?
Anthropic CEO Dario Amodei published an essay last week warning of existential AI risks, pledging to slow frontier development, and calling for limited antitrust exemptions for safety cooperation.
OpenAI global policy chief Chris Lehane said the company has been working with Anthropic and Google DeepMind on AI risks for several weeks, citing the aviation industry as a precedent.
In plain terms = the leading AI labs want to team up on safety, but under current law, competitors sitting down together can itself be a violation — so they need an official pass.
Is the DOJ's door open or closed?
Woodward said the government is watching "very closely," but stressed: no AI company has reached out to his office so far.
He made a public offer: "If anyone from any frontier lab wants to come talk about their concerns, we will take the meeting — but that request hasn't come."
This reflects a subtle signal: the DOJ hasn't shut the door, but it hasn't knocked either — the ball is in industry's court.
Why is the FTC pushing back?
FTC Chair Andrew Ferguson said he would be "highly vigilant" about exemption requests.
His core argument: AI companies seeking exemptions are effectively "seeking to erect barriers to entry" to shield themselves from challengers.
This means → in the FTC's view, "safety cooperation" and "ganging up against rivals" are separated by a thin line — and it leans toward seeing the latter.
Two agencies disagree — what happens next?
The DOJ and FTC share jurisdiction over competition enforcement. Either one's opposition can stall policy progress.
In plain terms = it's like a car with two steering wheels, one turning left and the other right — the car goes nowhere in the short term.
For the AI industry, a unified antitrust exemption is almost certainly off the table for now. Safety collaboration will continue to navigate a legal grey zone.
市场有风险，内容仅供研究参考，不构成投资建议。